The purposes of the study are (1) to determine the relations among ratings-made by parents, teachers, child care workers, and an independent observer--of the behavioral problems and competencies of children in residential and day psychiatric treatment; (2) to determine the relations between children's initial behavior patterns as reported by these sources and the children's progress in treatment. The initial ratings are made by parents, Hillcrest staff, and the independent observer at the beginning of the academic year and again in the late spring. Relations between initial behavior patterns--derived from ratings by the various sources--and children's progress--as measured by the follow-up ratings and indices of academic functioning--will be compared with similar data on out-patient samples to determine the prognostic validity and generality of the initial behavior patterns.